Determining the Appropriate Methodology for the Surveillance System

Objective IV Outline

Learning Objectives

Define the Injury events and data elements to be included in the system
Develop the data collection instrument and determine data collection frequency
Plan for Systemization, Maintenance and Data Security
Define Key positions

Introduction (will touch briefly on the following topics)
Discussion of what people want in their system. – Comprehensive or specific – i.e. suicide, domestic violence, etc.
Discuss ideal versus reality - tailored locally based on resources
Size and type of injury problem
Available data sources
Feasibility of getting information
Political aspects of getting information
Potential for defining interventions
Define the injury events and determine the data elements to include in the system
Case Definition (Pages 32-41)
Injury pyramid
Death (the tip)
Data Readily accessible
Cause of death from injury Consistently reported
Rare event
Not a good guide to ascertain the overall injury problem or medical consequences
Influenced by small numbers, especially population
Hospitalizations (below deaths)
When combined with mortality data, provides a much better picture of injury problem
Disability and health care costs can be better described
Access to data more difficult (privacy, HIPAA, manual records)
Incomplete or inconsistent coding of injury causation.
Data collection consumes human resources
ED visits (below Hospitalizations)
When combined with death and hospitalization data, helps provide the big picture
Small populations with limited injury mortality & hospitalization may benefit by casting the net wider to include ED visits
ED visits data can be useful for specialized study
Large number of cases may be difficult to manage
Access to data may be difficult – manual records
Inconsistent or incomplete identification of injury causation
Outpatient visits – Bottom of the pyramid
May be good for specialized injuries (sports related, eye injuries)
Might be good supplemental information, but in general not the place to start with injury surveillance
Difficult access (data highly protected by practitioners)
Summary
Don’t try to do everything at once
Start small with deaths and most severe injuries
Plan to expand or phase-in other levels of severity as your resources permit
You can work on prevention projects without knowing everything about every injury in your community
The use of e-codesin your system
E-codes – the value of e-codes
Allow the ability to identify tends
Allows the ability to describe the specific causes and contributing factors associated with an injury.
Limitations of e-code
Records not always coded
Miscoding or inconsistent coding
Poor Chart info results in non-specific e-code
Don’t always provide the desired specificity
Must stay apprised of updates
Reading an e-code and coding data -- exercise – Note: 30 minutes was allotted for this exercise in IHS Course. The examples below were taken from IHS Material. Using ICD-9 (or 10?) code book, participants are asked to code the following events.
Pedestrian run over by motor vehicle while walking in crosswalk at corner of Main Street and 3rd Avenue.
Unrestrained passenger of pickup truck ejected when tire blew, driver lost control, ran off road, and overturned.
36 year old female diagnosed with depression attempted suicide by hanging.
Boy admitted to hospital with BB gun shot to eye.
Intoxicated individual fell out of bed of stationary pickup truck.
Wife was assaulted and stabbed with knife by husband.
Severe allergic reaction due to scorpion bite.
Smoke inhalation from house fire that was caused by faulty wiring. There was no smoke detector in the home.
Neck injury due to diving and striking head against bottom of swimming pool.
Fall off bull during rodeo.

Case definition exercise page 40 – Time ?
Use the provided scenario to develop a case definition based on a real concern of someone in your group.
The primary data you intend to use is hospital ED records and patient medical charts
Considerations: person, place, time, intent, severity
Debrief points:
Do you think you will capture all injuries meeting case definition
What additional data sources should you consider
Group comments
Case definition chart page 42 -- chart compares case definitions from the public health perspective (using ICD-10 codes) vs. the law enforcement perspective.

Defining Variables
General Discussion
Keep the goal in mind – goal is intervention, prevention activity
Be mindful of ideal vs. reality
The simple form more likely to yield information, although it may not be everything you want to know.
Consider subsequent forms to gather more details
The variables you choose to include will depend on your locality, culture, etc. – rodeo injury related variables vs. snowmobile injury related variables
Name – often not available, not used or not collected. Use hospital case number or DOB in lieu of name. Privacy/HIPAA issues.
Age and sex –
Education Level – consider the need for the information. Will the education level impact your intervention activities
Employment Level – Same consideration as Education Level. Will it impact intervention.Why you might want it: Could help determine work related injuries.
BAC – important variable, but not always available. May have to settle for Alcohol involved (needs to be defined) and Alcohol related (needs to be defined)
Occupant Protection for Transportation – horses, skateboards, motorized vehicles, such as snow mobiles, boats – seat belts, helmets, life jackets, etc.
Time Related variable:
Date and time of injury event, ideally time the injury occurred. Realistically most of the time you only know the time of medical treatment. Choose one variant and go with it, even if it’s not the most accurate indication of when the injury occurred.
A drawback: in some cases people don’t seek medical treatment until much later. For example, A Navajo clinic was getting a lot of people coming in for fall related injuries on Monday morning, but the injuries were occurring over the weekend.
Place Related Variables:
Place where injury occurred, residence of injured person.
Injuries often occur when people travel to places for activities – hunting, fishing, drinking -- ideally you collect both. But sometimes the information is very general – such as the nearest village. Specifics are great, but if you can’t get them you work with what you have.
Variables related to the event circumstances
Relationship of victim to aggressor
Mechanism
Context
Criminal history of victim and/or aggressor
Example: Phase 1/Phase 2-- There’s value in collecting as much information as possible, but the more information you try to collect, the likely your form will be filled out accurately or at all. Some IHS areas have used a two phased approach. -- Phase 1 form used to collect basic information about injury, Phase 2 used to collect more detailed information based on the specific injury
Chart showing variables collected use forms supplied by group members as example of what variables are collected. Stress that each should customize forms to fit the needs of their community and available data. Give the example of white river having to re-do their entire system based on changes out in the real world.
You must define what you want in your system and that leads to the creation of a form.
System protocol examples – show what decisions you made in determining how to run the system.

Develop Data Collection Instrument and Determine Data Collection Frequency
Now must determine how design a form to capture the data/variables, we’ve decided to include.
Data Collection Planning – most data is record abstraction, gathering data from somebody, recording what you need. on-going surveillance –
Discussion of length – page 50
Keep it simple
Only include the data you need and then use analysis to answer case definition questions later.
The simpler the form, the greater its usefulness over a long period of time, the more likely you’ll be able to collect that data over a long time. There’s no right or wrong approach, but length is a consideration.
Keep the form easy on the eye
Decide whether or not to precode the form. Provide a precoded list of possible answers to a question as opposed to open-ended questions. When possible use numbers rather than letters. Numbers a re easier to process and less prone to error.,
Examples of forms from different area offices
Each surveillance system is unique and will have access to different data sources.
No such thing as a perfect system, find something that works for your needs
Pre-test - page 51
Pretesting done with a draft form of the instrument to allow changes to be made prior to a full investigation.
Pretesting helps identify if the questions and format are appropriate, clear, relevant, and result in the appropriate data. Help determine if the case definition is accurate. It’s done with a draft form of the instrument to allow changes to be made prior to a full investigation.
Not unusual to change a form two or three times or more before implementing it.
Data Collection Planning Summary
Decide what you want out of surveillance
Identify your case definition
Define your variables
Develop your form
Test it
Include discussion of when to change or abandon your form
Include discussion of quality control in general for the surveillance system
Frequency of Data Collection up to each locality to determine. Some considerations:
The protocol for your system,
Your needs
Your resources – human and financial
The requirements of the stakeholders
The magnitude of the injury problem in your area
Active versus passive collection. Active collection where a patient is sought out and interviewed about his or her injury is not done as a rule. Most information comes from police reports, ER logs, etc. Patients are not usually questioned because it’s not practical in most cases. Most surveillance in IC is passive in that the data was extracted from forms that were filled out by health care providers in the course of doing their job – for example death certificates or ER logs filled in by doctors. See the list of possible data sources in the appendix.
How confidentiality laws may impact your efforts –You should be familiar with HIPAA – the Health Insurance Portability and Accountability Act. This law, passed in 1996 safeguard’s an individual’s private health information (PHI) andit may impact your surveillance efforts. Consult with your local Institutional Review Board if your area has one or with the medical records coordinator at facility where you are gather data regarding your surveillance efforts and federal privacy laws. For more information on HIPAA visit or check the appendix for a list of classes.
Determine the type of Surveillance System -- there are several ways of setting up a surveillance system depending on the objectives to be met, budgetary considerations, the size of your staff, your locality. The Pan Am Health Organization describes several types of Surveillance systems some of which may not be appropriate for your needs
Universalsurveillance:Thetotalnumberofcasesoccurringwithinadefined populationisincludedinthesystem.Thispopulation-basedsurveillance accountsforallcasesthatoccur.Thisisthepreferredmethodofmonitoring theoccurrenceoffatalinjuriesbecauseratesofinjuriesandinjuryriskfactors canbecalculatedandgeneralizedtothepopulation. Most surveillance in Indian Country is an attempt to capture all data. The methods below may be done as a follow up to capture more data on specific injuries.
Surveillancebasedonsamplesofcases:Theinformationisobtainedfrom aportionofthetotalnumberofcasesorevents.Thesamplemustbe representativesothatinferencescanbemaderegardingallpossible casesoccurringinthepopulation.Thismethodcanbeusedtocollect informationaboutnonfatalinjuries or as follow up to collect more data on specific injuries.
Surveillancebasedonareviewofinstitutionalregistries:Institutional registriesarereviewedperiodicallytoanalyzeandidentifyvariablesof interest.Whenusingthismethod,itisimportanttoproperlyidentifythe institutionsandthesourceswithininstitutions,suchasclinicaland emergencyrecords,hospitaldischarges,orcomplaintsfiledwithpoliceor familywelfareinstitutions.Itisusefulformonitoringspecificinjuries.
Survey-basedsurveillance:Informationisobtainedthroughquestionnaires focusedonaspecifictopic,withina predefinedperiodoftime,andat predefinedintervals.IntheUnitedStates,forexample,self-reportedseat beltandsafetyseatuseismeasuredatthestatelevelbyhouseholdsurveys conductedfortheBehavioralRiskFactorSurveillanceSystem(BRFSS),by school-basedsurveysconductedfortheYouthRiskBehaviorSurveillance System(YRBSS),andbydirectobservationofpassengervehicleoccupantsfor theNationalOccupantProtectionUseSurvey.11
Sentinelsurveillance:Oneormoreinstitutionsarechosentomonitor trends,targetsurveillanceactivities,andsuggestpreventiveinterventions. Ingeneral,surveillancesystemsofthistypearenotrepresentativeofthe population,butareusefulforcallingspecialattentiontorisksituationsandthusfulfillakeyfunctionforinjurypreventiondecision-making. Oneexampleofthistypeofsurveillanceistheapproachtakenbychild deathreviewteams,whichgatherandanalyzedataonthecircumstances surroundingallcausesofchilddeaths.Sentinelsurveillancesystems complementothersourcesofinformationforinjuryprevention.
Plan for Systemization, Maintenance and Data Security – Quality control issues and protocol
If you don’t have an electronic database set up, seek the assistance of a programmer, an epidemiologist or a statistician. Epi Info 2002 is free software available through the CDC. For an overview of the software and installation and use instructions visit
Data maintenance should focus on the following
Reducing errors that may be introduced through flaws in the design.
Improving the systems scope and services through routine maintenance, emergency maintenance and requests for special reports. Should be based on local needs. consider some of the following
Backing up data and system files according an established schedule.
Maintaining records in a secure environment
Requiring requests for emergency maintenance to be in writing and entered into a log
Assigning priorities for special requests on the basis of urgency of need and time and resources required for fulfillment
Institutionalizing routine maintenance
Documenting maintenance that is conducted
Safeguarding your system. Should be based on local circumstances.
Consider limiting access to one person
Consider installing the database on two computers. A primary computer and one reserved for testing changes to the system. The second computer can also serve as a backup in case the first one fails.
Consider keeping a second copy of the database off site. Routine updates of the offsite copy must be done.
Threats to a database
Human error
Mechanical failure
Malicious damage
Cyber Crime
Invasion of privacy
Computer viruses
Protocol – Successful surveillance systems will have clearly written protocol Include examples provided by Workgroup. Will vary with needs
Include procedures for obtaining and securing data
Maintenance
Rules for data storage
Rules for password protection and passwords
Define staff and key positions necessary
Ideally a surveillance system will have a core staff, both part time and full time, that would include program manager, data manager, research analyst and coordinator. This is seldom if ever possible in Indian Country. In Indian Country it is more like that your staff will consist of one or two people who will fulfill all of the functions of the system. Some of the key staff functions include:
Coordinate system activities
Establish contact with data sources and stakeholders
Data entry
Quality Control
Analysis
Preparation of Reports
Advisory Board/Coalition – Just as with coalitions mentioned early an advisory board if possible could be helpful for offering technical advice, strategic planning and support for the surveillance system, but the reality is that often you will be working with just a few people. Whether your advisors are a board or just a few people, it may be necessary from time to time to seek a group or an individual’s assistance with the following
Obtaining the data necessary for the injury surveillance system
Review and Advice on policy and procedures for data collection, linkage, publications, and mechanisms for implementing a reporting system
Identifying the best use of data
Strategizing about how to remove obstacles and inefficiencies
Providing speaking opportunities with professional organizations
Obtaining data sharing agreements
Showing broad, high-level support for the system.
Getting local approval to start a surveillance system
Navigating tribal politics or resistance to surveillance, data collection or data sharing
Summary

1 / 7/10/12